Download Differential expression of fibroblast growth factor receptors

Differential Expression of Fibroblast Growth Factor
Receptors During Rat Lens Morphogenesis and Growth
Robbert U. de Iongh, Frank J. Lovicu, Coral G. Chamberlain, and John W. McAvoy
Purpose. Fibroblast growth factors (FGF) play important roles in the developmental biology
of the lens. Recently, it was shown that the expression of one of the FGF receptors, FGFR1
(fig; fibroblast growth factor receptor 1), was closely associated with the onset of lens fiber
differentiation. In this study, the expression patterns of three other members of the FGF
receptor family were analyzed and compared.
Methods. The expression patterns of FGFR2 (bek and keratinocyte growth factor receptor
[KGFR] variants) and FGFR3 were analyzed by in situ hybridization during embryonic and
postnatal lens development.
Results. In the ocular primordia, both FGFR2 variants were detected on embryonic day 12
(E12) and FGFR3 was detected on E14. From E16 to E20, distinct spatial expression patterns
became evident within the lens; FGFR3 showed an anteroposterior increase in expression, with
strongest expression in the outer cortical fibers. In contrast, bek showed uniform expression
throughout the lens epithelium (including the central and germinative zones) and the transitional zone, with a subsequent decline in maturing fibers. The KGFR variant of FGFR2 showed
strongest expression in the earlyfibersof the transitional zone; its expression in the epithelium
was weaker in the germinative zone of embryonic and neonatal rats. There was an age-related
decline in expression of FGFRs after birth—an effect that was more marked for FGFR3 than
for the FGFR2 variants.
Conclusions. Combined with those in a previous study, these results indicate that the FGFR1,
bek, KGFR, and FGFR3 genes exhibit different, yet overlapping, patterns of expression throughout lens development and differentiation. The distinct spatiotemporal patterns of expression
of FGF receptors may play an important role in regulating anteroposterior patterns of lens
cell behavior. Invest Ophthalmol Vis Sci. 1997;38:1688-1699.
Jr ibroblast growth factors (FGFs) constitute a family of
at least 10 structurally related polypeptides, which are
highly conserved between species and induce a wide
range of responses in various cell and tissue types, including proliferation, differentiation, matrix deposition, and
cell migration. The FGFs show distinct spatial and temporal expression patterns in embryos and adults and are
involved in many key developmental processes, including
From the Department of Anatomy and Histology and Institute for Biomedical
Research (FB), The University of Sydney, Australia.
Supported by grant RO1 EY03177 from the National Eye Institute, US Department
of Health and Human Services, Public Health Service; by a grant from the National
Health and Medical Research Council (NHMRC), Australia; and by an NHMRC
Biomedical Research Scholarship and a traveling scholarship from the Faculty of
Medicine (Rdel) and a Postdoctoral Research Fellowship from the Medical
Foundation (FJL), University of Sydney, Australia.
Submitted for publication November 20, 1996; revised March 27, 1997; accepted
March 31, 1997.
Proprietary interest category: N.
Reprint requests: fohn W. McAvoy, Department of Anatomy and Histology (F13),
University of Sydney, Sydney NSW, Australia 2006.
1688
determination of the anteroposterior axis and induction
of mesoderm during early embryonic development.'
Fibroblast growth factors play a pivotal role in lens
differentiation.2 The differentiated lens has a highly ordered cellular architecture. It is composed of two distinct
forms of lens cell: elongated fibers, aligned in an anteroposterior axis, make up the bulk of the lens; and a monolayer of cuboidal epithelial cells covers the anterior surface of the fibers. The lens grows by proliferation of epithelial cells in the germinative zone of the lens, just
anterior to the lens equator; and progeny of these divisions move posteriorly into the transitional zone of the
lens where they elongate and differentiate into fibers.
This process continues to add fiber cells to the lens mass
throughout life, so that the position of a fiber cell within
the fiber mass (from outer cortex to nucleus) reflects its
state of differentiation. Results of previous studies in vitro
have shown that FGF induces lens epithelial cells to un-
Investigative Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9
Copyright © Association for Research in Vision and Ophthalmology
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 08/03/2017
FGF Receptor Expression in Lens Development
dergo proliferation, migration, and fiber differentiation
in a progressive dose-dependent manner.3 There is increasing evidence that an anteroposterior gradient of FGF
stimulation plays an important role in regulating lens
polarity and growth patterns.2 This hypothesis is supported by results of recent transgenic studies, which indicate that overexpression and inappropriate secretion of
FGFs in the lens induce differentiation of the anterior
epithelium and abolish lens polarity.4"7
The FGFs bind two distinct types of cell-surface receptors, high affinity tyrosine kinase receptors (FGFR)8 and
lower affinity heparan sulfate proteoglycans, both of
which are required for biologic activities of FGF.910 There
are at least four FGFR genes, and alternative splicing of
their messenger RNAs (mRNAs) gives rise to several variants that differ in their affinities for members of the FGF
family.811 Ligand binding to the extracellular domain of
the FGFR induces receptor dimerization, which results in
transphosphorylation of tyrosine (s) in the intracellular
tyrosine kinase domains of the receptors in the dimer.12
Activation of the kinase domain permits binding and
phosphorylation of intracellular signaling proteins and
activation of specific intracellular signaling pathways. The
importance of FGF signaling through FGFRs for lens development has been demonstrated by findings in
transgenic studies in which a dominant-negative FGF
receptor was expressed in the lens fibers to inhibit FGF
signaling.1314 These findings showed that lens fibers are
dependent on FGF signaling through FGFRs for normal
differentiation and survival.
Recently, results of detailed studies of the spatiotemporal expression patterns of FGFRl during lens development produced results that established that a high level
of expression of FGFRl was associated with the onset of
lens fiber differentiation.15 In other studies, results have
indicated that lens cells also express variants of FGFR216
and FGFR317; but, because these studies were restricted
to a single early-stage embryo, information on the spatial
expression patterns in the lens was limited, and there
was no information on the temporal expression patterns
during lens development.
This report presents a detailed, in situ hybridization
analysis of the expression patterns of FGFR2 and FGFR3
throughout embryonic and postnatal lens development.
Specific probes for bek and keratinocyte growth factor
receptor (KGFR) allowed detailed analysis of the expression of these alternatively spliced variants of FGFR2. The
results of this study establish that mRNAs for these FGFRs
have unique, yet overlapping, patterns of expression during lens development and that they are present in different proportions in the different zones of cellular activity.
METHODS
Animal and Tissue Preparation
All procedures involving animals were in accordance
with the National Health and Medical Research Coun-
1689
cil (Australia) guidelines, the National Institutes of
Health Guide for the Care and Use of Laboratory
Animals (NIH; Bethesda, MD) and the ARVO Statement for the Use of Animals in Ophthalmic and Vision
Research.
Embryos at various stages of gestation (embryonic
days 11 to 20; Ell to E20) and eyes from neonatal
(postnatal day three; P3), weanling (P21), and adult
(PI00) rats were immersed in Tissue-Tek OCT compound (Miles, Elkhart, IN) and frozen in iso-pentane
cooled by liquid nitrogen. Specimens were stored in
liquid nitrogen until sectioned.
Complementary DNA Probes
Two complementary DNAs (cDNAs) for the alternatively spliced forms of murine FGFR2 (bek and KGFR:
Orr-Urtreger et al16) and coding for the variable regions of the third immunoglobulin domain were obtained from Dr. M. Bedford and Dr. P. Lonai (Weizmann Institute of Science, Rehovot, Israel). The murine bek cDNA was a 119-bp fragment (PPum I-Eco
RV), which was proved by sequence analysis (BLAST
analysis at the National Center for Biotechnology Information, NIH) to be 99% homologous with the published rat bek cDNA.18 The murine KGFR cDNA was a
160-bp fragment (PPum I-Hae II), of which the first
158 bp were 100% homologous with the published rat
KGFR cDNA.18 Both variants of FGFR2 were subcloned into pBluescript (Stratagene, Lajolla, CA). A
487-bp cDNA encoding the complete third immunoglobulin domain, including exon Illb, of murine
FGFR3, subcloned into pBluescript KS+, was obtained
from Dr. L.T. Williams (University of California, San
Francisco). The first 291 bp and the last 44 bp of this
cDNA code for common regions in the Illb and IIIc
exon forms of FGFR3,19 whereas nucleotides 292 to
443 are specific for exon Illb only. In sequence analysis, this sequence proved to have regions with high
levels of homology (99%) with human FGFR3 sequences but only moderate homology with rat FGFRl
(72%), FGFR2 (72%), or FGFR4 (73%). RNA probes
transcribed from these cDNAs were used for in situ
hybridization experiments under high-stringency conditions, as described previously.15"2021
In Situ Hybridization
In situ hybridization was performed on paraformaldehyde-fixed frozen sections of rat ocular tissues, using
SP6, T3, or T7 RNA polymerase-derived RNA probes,
labeled with 35S-UTP (Amersham, Sydney, Australia).
After hybridization and final high-stringency washing
(0.1 X SSPE at 65°C; 1 X SSPE = 15 mM NaCl, 1 mM
NaH2PO4, 1 mM EDTA), sections were dehydrated
and exposed for 5 days to autoradiograph film (/3-Max
Hyperfilm, Amersham), which was developed according to manufacturer's instructions. Slides were
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 08/03/2017
1690
Investigative Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9
then coated with NTB-2 emulsion (Kodak, Sydney,
Australia) and stored in light-tight boxes with desiccant at 4°C for 3 to 5 weeks, after which they were
developed (D-19, Kodak), rinsed, and stained with hematoxylin. Sections were photographed using a Leitz
Dialux 20 microscope (Wetzlar, Germany) and 400
ASA film (T-Max, Kodak) processed according to manufacturer's instructions.
Image Analysis of Fibroblast Growth FactorReceptor Hybridization Signal
Image analysis was used to compare the density of
FGFR expression of different regions of the fetal (E20)
lens, as previously described.15 Briefly, dark-field images were captured directly from slides by video camera (slow scan, Dage/MTI, Michigan City, IN) and the
optical density of hybridization signals was measured
in six distinct regions (Figure 3) of lens sections, using
a Tracor Northern Image Analysis system (Tracor
Northern, Middleton, WI, USA). For each region, the
perinuclear area that contained the specific signal was
delineated, and the optical density of that region was
measured. The regions measured were consistent for
each of the different FGFR probes analyzed. To control for variability in hybridization and autoradiography conditions between experiments, data for each
experiment were first corrected for nonspecific background and then normalized for the mean count for
region I. Values obtained were compared using oneway analysis of variance and Student's t-test. This approach allowed comparison of the relative expression
of each of the FGFR mRNAs in the various regions
of the lens. However, it does not necessarily reflect
changes in cellular expression for a given receptor, as
dramatic changes in cell shape and size occur as lens
cells differentiate into fibers.
The changes in expression of each FGFR with age
were compared by laser densitometry. As described
previously,15 sections of whole eyes from P3, P21, and
PI00 rats, which had been hybridized with the same
probe under identical conditions, were exposed to an
autoradiograph film (/?-Max Hyperfilm, Amersham)
for 5 days. The film was scanned using a HeNe laser
densitometer (Molecular Dynamics, Sunnyvale, CA),
and the density of signals in region III of P3, P21,
and P100 lenses was quantified using image analysis
software (ImageQuant; Molecular Dynamics). After
correction for background, the data were compared
using one-way analysis of variance and Student's Rest,
as described.
RESULTS
Transcripts for two alternatively spliced variants of
FGFR2 {bek and KGFR) and for FGFR3 were detected
at various stages of lens development. At no stage were
distinct signals detected with the respective sense
probes (not shown).
No distinct hybridization signals for bek, KGFR, or
FGFR3 were detected at Ell (data not shown). At El2,
there was detectable expression of bek transcripts in
the lens pit and surrounding mesenchyme (Fig. ID).
FIGURE l. Expression of FGFR during lens morphogenesis. Expression of bek (D to F), KGFR
(G to I) and FGFR3 (J to L) in E12 (A,D,GJ), E14 (B,E,H,K), and E16 (C,F,I,L) embryos.
A, B, and C are hematoxylin-stained sagittal sections through the ocular primordia, as shown
in D, E, and F by dark-field microscopy. The section shown in B and E is slightly tangential
to the sagittal plane. At El 2, distinct expression for bek (D) was detected in the lens pit (lp),
anterior parts of the optic cup (oc), diencephalon (d), and undifferendated extraocular
mesenchyme (m), whereas only weak KGFR expression was found in ectoderm in the area
around the lens pit {arrowheads, G), and FGFR3 was undetectable (J). At E14, the lens is
comprised of an epidielium {arrowhead) and a differentiating fiber mass (If). Bek (E) and
KGFR (H) were expressed in the epithelium and fibers of the lens vesicle, whereas FGFR3
(K) was expressed only in die fibers. All three receptors were detected in presumptive
choroid sclera (cs) and in the cartilaginous mesenchymal condensations representing bone
precursors {solid arroius), albeit weakly for FGFR3 (K). KGFR and, to a lesser extent, FGFR3
were strongly expressed in the invaginating ectoderm of the presumptive eyelids {curved
arroius, H,K). In the lens (1) at E16, bek (F) and KGFR (I) were expressed in lens epithelium
{small arrowheads) and in the early fibers at the equator {large arrowheads), whereas FGFR3
(L) was expressed in all lens cells but was particularly strong in the elongating fibers. Bek
transcripts were also found in the developing cornea (c), ganglion cell layer of the neural
retina (nr), the peripheral redna (pr; presumptive ciliary body and iris), and in the choroid
sclera (cs). KGFR expression was detected in the peripheral retina, the choroid sclera and
the invaginating ectoderm of the presumptive eyelids {curved arroiu). FGFR3 was also detected
in the eyelid ectoderm {curved arroiu) and developing bone of the skull {solid arrow). Scale
bar = 50 //m except for A, D, G, and J, for which scale bar = 20 /xm.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 08/03/2017
FGF Receptor Expression in Lens Development
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 08/03/2017
1691
1692
Investigative Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9
Although little or no signal was detected in the optic
cup, it was detectable in the diencephalon. Weak signals for KGFR transcripts were only detected in the
ectoderm in the area around the lens pit (Fig. 1G).
No discernible signals for KGFR were present in the
lens pit or optic cup. Distinct signaling for FGFR3 was
not detected in the eye primordia at this stage (Fig.
u>-
At El4, strong signals for bek were detected in the
epithelial cells and in early fibers of the lens vesicle
(Fig. IE). Signals were also detected in the peripheral
optic cup, in condensing mesenchyme of the choroid
sclera and in presumptive bone (Fig. IE), but not in
the central optic cup or in regions of undifferentiated
mesenchyme (Fig. IE). With the KGFR probe, signals
were detected in the lens vesicle, particularly in the
early-elongating primary fibers (Fig. 1H). Other tissues that showed signals for KGFR included condensing mesenchyme of the choroid sclera, presumptive
bone (Fig. 1H) and the imaginations of the ectoderm
that give rise to the eyelids and conjunctival epithelium (Fig. 1H). At this stage, transcripts for FGFR3
werefirstdetected in the elongated primary lens fibers
(Fig. IK), but not in the epithelium. Strong signals
were detected in the invaginating ectoderm of the
presumptive eyelids (Fig. IK). Weak signals for FGFR3
were detectable in the mesenchyme of the choroid
sclera and in the presumptive bone (Fig. IK).
With further differentiation of die lens at E16,
the hybridization signals for the three FGFR probes
showed more distinct spatial expression patterns.
Strong signals for bek transcripts were detected in the
anterior epithelium and equatorial regions of the lens
(Fig. IF). Strong signals were also present in the cornea and regions of the peripheral retina destined to
form the ciliary body and iris (Fig. IF). Weaker signals
were found in the developing ganglion cell layer and
in the choroid sclera (Fig. IF). For KGFR, strong signals were found in the equatorial lens, presumptive
ciliai'y body, and iris, with strongest signals in the ectodermal invaginations of the eyelids. Weaker signals
were found in the anterior lens epithelium and in the
choroid sclera (Fig. II). Transcripts of FGFR3 were
found throughout the E16 lens but were most strongly
localized in the elongating lens fibers. Signals were
also detected in the eyelid ectoderm imagination and
in developing bone (Fig. 1L). In the elongating fibers,
the signal for all three FGFR probes was predominantly perinuclear (Figs. IF, II, 1L).
At E20, distinct spatial expression patterns for the
three FGFRs in the lens were well established. In the
lens, a uniform signal was detected for bek transcripts
from the central epithelium, through the germinative
zone, to the transitional zone, which is located posterior to the lens equator (Fig. 2B). The signal diminished in die cortical fibers that had undergone further
FIGURE 2.
Expression of FGFRs in the lens at £20. (A) Hematoxylin-slained section of E20 eye showing lens epithelium
(le) lens fibers (If), cornea (c), iris (i), ciliary body (cb),
neural retina (nr), choroid sclera (cs), eyelid epidermis (e),
eyelid suture (s), and eyelid mesenchyme (em). Dark-field
micrographs show expression of bek (B), KGFR (C) and
FGFR3 (D) in the E20 lens. (B) Transcripts for bek at E20
showed a uniform distribution throughout the epithelium,
(including the central and germinative zones) and transitional zone. Signals decreased progressively in the inner and
outer cortical fibers and in the mature fibers (asterisk). See
Figure 3A for definition of lens regions. (C) Signals for
KGFR appeared uniform in the anterior central lens epithelium, decreased slightly in the germinative zone, and then
increased in the transitional zone. No significant signals
above background were detected in the maturefibers(asterisk). (D) Signals for FGFR3 was detected anteriorly in the
central epithelium and increased in the germinative and
transitional zones. Strongest signals were detected in the
cortical fibers (arroioheads). No significant signals above
background were detected in the mature fibers (asterisk).
Scale bar = 50 /im.
elongation and differentiation and was virtually absent
in the most mature fibers in the center of the lens
(Fig. 2B). In contrast, the signals for KGFR and FGFR3
showed an anteroposterior increase widi weaker signals in the central epithelium than in the more posterior regions (Figs. 2C, 2D). Signals for FGFR3 were
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 08/03/2017
FGF Receptor Expression in Lens Development
particularly strong in the more mature fibers of the
outer cortex of the lens (Fig. 2D). For KGFR and
FGFR3, the signals diminished with further differentiation of lens fibers in deeper regions of the lens (Figs.
2C, 2D). As in E16 lenses, the signal for all three
FGFR probes was predominantly perinuclear: This was
particularly evident in the elongated fibers of the lens
cortex (Figs. 2B, 2C, 2D).
Outside the lens, strong signals for bek were detected in the ciliary body, the iris, and the epidermis
and hair follicles of the eyelids, with weaker signals in
the corneal stroma, neural retina, and mesenchyme
of the eyelid and choroid sclera. For KGFR, strong
signals were detected in the basal layer of the eyelid
epidermis (including the eyelid suture and hair follicles) and were continuous with signals in the conjunctival and corneal epithelia; relatively weak signals were
detected in the ciliary body and iris. For FGFR3, although signals were not detected in the surface epidermis, strong signals were present in the epidermal
cells at the eyelid sutures and in the conjunctival and
corneal epithelia.
In the lens, quantitative analysis of the hybridization signals confirmed that the signal density for bek
was uniform from the anterior epithelium to the germinative zone (Fig. 3B), whereas signals for KGFR and
FGFR3 increased in the more posterior regions of the
lens and subsequently decreased in fibers undergoing
later stages of fiber differentiation (Figs. 3C, 3D). Peak
expression for KGFR occurred in region III (transitional zone) and for FGFR3 in region IV (early cortical
fibers). Interestingly, the signal for KGFR showed a
small but significant decrease in region II (germinative zone) of E20 lenses (Fig. 3C). A slight decrease
in region II was also observed in P3 (P < 0.05), but
not in P21 (see Fig. 4E) or in P100 lenses.
In postnatal rats, the patterns of FGFR expression
established at E20 persisted, with bek showing a uniform distribution in the lens epithelium from anterior
to posterior regions, whereas KGFR and FGFR3 signals
increased in the transitional zone and outer cortex,
respectively (Fig. 4). This anteroposterior increase was
most pronounced for FGFR3 (Figs. 4C, 4F). With increasing postnatal age (P21 and P100), the signal for
KGFR and FGFR3 in the cortical fibers became more
restricted, and strongest expression for FGFR3 was
often found in the posterior parts of the transitional
zone, rather than in the cortical fibers (Fig. 4F).
Signals for bek and KGFR were detectable in the
ciliary body, iris, and cornea at P3 (Figs. 4A, 4B) and
P21 (Figs. 4D, 4E), but no distinct expression of
FGFR3 was detected outside the lens (Fig. 4). In the
neural retina at P21, bek, and KGFR showed essentially
similar patterns of expression, with weak signals present in the ganglion cell and inner nuclear layers and
along the outer edge of the outer nuclear layer (Figs.
1693
5B, 5C). Distinct signals for bek, but not for KGFR,
were present in the pigmented epithelium and weak
signals for both were detected in the choroid sclera.
Densitometric measurement of signal intensity in
lens region III from autoradiographs of P3, P21, and
PI00 eye sections showed that there were age-related
declines in expression for all three receptors, but this
was most evident for FGFR3 (Fig. 6). For bek and KGFR
forms of FGFR2, there were significant declines in
signals from days P3 to P21 (P < 0.005), but not from
days P21 to P100 of postnatal development. For
FGFR3, there was a progressive, significant decline in
expression during both periods of postnatal development (P= 0.0001).
DISCUSSION
The findings in our investigation have established that
two-splice variants of FGFR2 (bek and KGFR) and
FGFR3 are each expressed in different patterns in ocular tissues, including the lens, throughout embryonic
and postnatal life in the rat. Results of previous studies
from this laboratory showed that FGFR1 also has distinct spatiotemporal expression patterns during rat
lens development.15 These findings are consistent with
those in studies that show that FGFR genes and their
isoforms have different patterns of expression in other
tissues during embryonic and adult life.22
During rat lens morphogenesis, FGFR1 is the first
FGF receptor to appear. At Ell, it becomes detectable
in the ocular anlage, with distinct expression in presumptive lens ectoderm and optic vesicle.15 Messenger
RNAs for both members of the FGFR2 family (bek and
KGFR) only become clearly detectable at El2: Bek is
expressed in the ectoderm and lens pit, whereas KGFR
is expressed only in the ectoderm in the area of the
forming eye. FGFR3 appears later at E14, with formation of the primary lens fibers in the lens vesicle. During these stages of development, each of these receptors shows distinct patterns of expression in various
ocular tissues.
Once the lens has acquired its distinct polarity,
differences in the spatial expression patterns of FGF
receptors become more evident. FGFR1 is expressed
weakly in the anterior epithelium, increases toward
the lens equator in the germinative zone, and reaches
a maximum in the transitional zone of the lens (region
III) where fiber differentiation begins.15 The two variants of FGFR2 show strikingly different patterns of
expression. Bek is uniformly expressed throughout the
lens epithelium (including the central and germinative zones) and transitional zone, with a subsequent
decline in the maturing fibers; KGFR expression is
detectable in the anterior central epithelium, but
tends to decline in the germinative zone, at least in
younger animals. Strongest expression occurs in the
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 08/03/2017
1694
Investigative Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9
FIGURE 3. Quantitative analysis of FGFR expression in different regions of the E20 lens. (A) Micrographs of the E20
lens were divided into six regions for image analysis: I, the
anterior epithelium; II, the germinative zone; III, the transitional zone; IV, the outer cortical fibers; V, the inner cortex;
and VI, the mature fibers. The density of hybridization signals for bek (B), KGFR (C), and FGFR3 (D) were assessed.
Values were normalized for the mean signal for each probe
in lens region I. Each bar represents the mean ± SEM of
determinations from 13 to 17 separate sections on 4 to 7
separate slides. (B) For bek, the same level of signal density
was found in regions I to III, but a progressive decline in
signal occurred in regions IV to VI (P < 0.05). (C) For
KGFR, expression density was lower in region II than in
regions I, III, and TV (P< 0.05), and maximal signal density
was observed in region III, with a significant decline in regions IV to VI. (D) FGFR3 expression showed a progressive
increase from region I to region IV, with almost a twofold
increase between regions I and TV (P < 0.005) and subsequent decrease between regions IV and VI (P < 0.0001).
0.0
i
II
in
iv
v
vi
Lens regions
early fibers of the transitional zone. Similar to FGFR1,
FGFR3 shows an anteroposterior increase in expression; however, strongest expression occurs in the
outer cortical fibers. Thus peak expression of FGFR3
occurs later in the fiber differentiation process than
does peak expression of FGFR1. For all probes, only
low levels of signal are present in the more mature
nuclear fibers. Notably, FGFR3 is more highly expressed in the lens than in other ocular tissues and,
with postnatal development, is almost exclusively expressed in the lens.
It is of interest that all four FGFR probes studied
have shown a predominandy perinuclear localization
of FGFR transcripts, particularly evident in elongating
and differentiating fibers of the lens cortex. Because
FGFRs are membrane proteins, their actively translated mRNAs would be expected to be associated with
the rough, endoplasmic reticulum of the cell that, in
chick and rat lens cells, forms a dense meshwork
around the nuclei.23 Thus, perinuclear localization
may reflect active translation of the FGFRs. Consistent
with this hypothesis is the finding that there is strong
correlation between FGFR1 mRNA and protein expression during lens development.15 However, perinuclear localization may also indicate that these mRNAs
have short half-lives.
A quantitative in situ hybridization approach, using image analysis techniques, confirmed tfiat the different FGFRs are expressed in distinct, yet overlapping, patterns in the developing and maturing lens.
Results of additional studies show that these expression patterns are distinct from that of glyceraldehyde3-phosphate dehydrogenase (GAPDH; de Iongh, unpublished data), which is generally considered to be
a "house-keeping" gene. The pattern of GAPDH is
uniform throughout the anterior epithelium and region III, with a slight rise in the germinative zone
(region II). Distinct from the FGFR genes, the expression of GAPDH decreases more rapidly after region
III and is negligible in regions V and VI.
The distinct spatial patterns of FGFR expression
are suggestive of their involvement in regulating
spatial patterns of lens cell behavior. For example,
FGFR1 and FGFR3 show major increases in expression at various stages of fiber differentiation,
whereas KGFR expression declines in the germinative zone. In lens epithelial explants, FGF induces
different responses at different concentrations; proliferation, migration, and fiber differentiation are
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 08/03/2017
FGF Receptor Expression in Lens Development
1695
FIGURE 4.
Expression of bek, KGFR, and FGFR3 in the postnatal lens. Sagittal sections of eyes
from P3 (A to C) and P21 (D to F) rats. For bek (A, D), there was uniform expression in
the lens epithelium, including the central epithelium (arrowhead), germinative (open arroio),
and transitional (solid arroio) zones, in P3 and P21 lenses. Signals diminished beyond this
region, and no significant signals above background were detected in the mature fibers
(asterisk). Strong signals were also evident in the ciliary body (cb) and iris (i), with weaker
signals in the cornea (c). (B) KGFR signals were detected anteriorly in the P3 central
lens epithelium (arrowhead), diminished slightly in the germinative zone (open arrow), but
increased in the transitional zone (solid arrow). Signals decreased in the cortical region, and
no significant signals above background were detected in the central lens (asterisk). Signals
were also detected in the ciliary body <cb) and iris (i). (E) AT P21, uniformly weak signals
for KGFR were detected throughout the lens epithelium. Signals were also detected in the
ciliary body (cb), iris (i), and cornea (c). For FGFR3 (C,F), expression was detected only
in the lens. It was weak in the anterior epithelium (arrowhead) but increasingly strong in
the germinative (open arrow) and transitional (solid arrow) zones. Signals were also strong in
the outer cortical fibers, particularly at P3 and were diminished in the inner cortical region,
with no signals above background evident in the mature fibers (asterisk). There is a refraction
artifact at the outer edge of the cornea in B, E, and F. Scale bar = 50 fj.ni.
induced in a progressive, dose-dependent manner.3
These cellular behaviors occur in the same sequence
in the lens in an anteroposterior pattern (that is,
from region I to region VI; Fig. 3).
The underlying mechanism that enables lens cells
to vary their response to FGF is not clear at present;
however, it is possible that their primary response may
depend on the predominant form(s) of FGFRs present on the lens cells, either as homo- or heterodimers.
For example, the patterns of FGFR1 and FGFR3 expression suggest that stimulation of these receptors
may be associated with events in fiber differentiation.
In contrast, bek may have a more general role in epithelial maintenance, proliferation, and early fiber differ-
entiation, whereas increased KGFR expression just beyond the germinative zone may be linked with withdrawal from the cell cycle and commencement of fiber
differentiation. The suggestion that different receptors may be involved in influencing the primary response of lens cells to FGF is supported by findings
in studies of FGFR signaling pathways in other cellular
systems. These indicate that activation of different
FGFRs by FGF can induce different patterns of intracellular protein phosphorylation and mediate different cellular responses.24'25
The temporal expression patterns for FGFRs
throughout development are also consistent with the
above suggestion that differences in cellular responses
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 08/03/2017
1696
Investigative Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9
FIGURE 5. Expression of FGFRs in the P21 retina. (A) Hematoxylin-stained section of P21
retina showing the cellular layers of the retina: ganglion cell layer (gel), inner nuclear layer
(inl), outer nuclear layer (onl), retinal pigmented epithelium (rpe), and the choroid sclera
(cs). Dark-field micrographs show expression of bek (B), KGFR (C), and FGFR3 (D). Bek
and KGFR probes showed similar patterns of expression with signals detectable in the
ganglion cell and inner nuclear layers. Most of the outer nuclear layer was not labeled but,
in each case, distinct signals were detected along the outer edge of this layer, corresponding
to the outer segments of the photoreceptors. Distinct signals for bek but not for KGFR were
found in the pigmented epithelium, and weak signals for both were found in the choroid
sclera. Signals for FGFR3 were not detected in the neural retina. There is a prominent
refraction artifact at the outer edge of the retina in C and D. Scale bar = 20 fj,m.
to FGF are related to differences in cellular populations of FGFRs. During embryonic development, there
is a distinct sequence of FGFR expression: first FGFR1,
then FGFR2, andfinally,FGFR3. Furthermore, FGFR1
and FGFR3, which are consistendy expressed most
strongly during periods of active fiber differentiation,
show a marked age-related, postnatal decline in expression (Fig. 6),15 similar to the documented decline
in lens fiber differentiation with age.26"28 In contrast,
bek and KGFR show a less marked decline in expression with age, indicating a role more in line with cell
maintenance than with fiber differentiation. Lens fibers are clearly dependent on FGF for survival and
differentiation, as indicated by findings in recent
transgenic studies in which a dominant negative FGFR
was overexpressed in the mouse lens (see introduction).1314 However, because the dominant-negative
construct inhibits all FGF receptors,29'30 these data do
not provide information about whether multiple or
individual FGFRs are involved in eliciting specific cellular responses.
Recendy, mutations of three of the FGFR genes
have been identified as responsible for several autosomal dominant human skeletal disorders.31 These mu-
tations are associated with unique skeletal phenorypes
that can be correlated with the expression patterns of
the FGFRs but cannot be explained simply in terms
of loss of gene function.31"33 For instance, FGFR gene
knockout studies in transgenic mice show that heterozygous animals have normal "wild-type" phenotype
whereas homozygous animals either are not viable embryonically, as in the case of FGFR1M'35 or have a phenotype distinct from the human syndromes, as in the
case of FGFR3.33 Comparison of the different patterns
of bone growth in the homozygous mouse FGFR3
gene knockout and the human FGFR3 mutant phenotype (achondroplasia) led Deng et al33 to conclude
that in the human syndrome there is a gain rather
than a loss of function: the mutant FGFR3 is constitutively active and, during normal terminal bone differentiation, FGFR3 may have a negative regulatory role
by inhibiting proliferation and terminal differentiation of chondrocytes. Because peak expression of
FGFR3 occurs in the lens in terminally differentiating
fibers, it is possible that it has an analogous, negative
regulatory role during terminal differentiation of lens
fibers.
Characteristic features of the human craniosy-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 08/03/2017
FGF Receptor Expression in Lens Development
A
bek
P3
P21
P100
Age of rat
6. Age-related decline of FGFR expression in the
lens. Sections of whole eyes from rats at P3, P21, and P100
were hybridized with probes for bek (A), KGFR (B), and
FGFR3 (C) under identical conditions and exposed to the
same autoradiographic film for 5 days. The density of the
hybridization signal in region III of lenses (Fig. 3) was measured by laser densitometry. Each column represents the
mean ± SEM of measurements from 5 to 8 lenses. Density
of hybridization signal for bek and KGFR probes declined
significantly between P3 and P21 (P < 0.005) but not between P21 and P100. Signal density for FGFR3 declined
progressively from P3 to P100 (P < 0.005).
FIGURE
nostosis syndromes, which result from mutations of
FGFR1, FGFR2, or both, include ocular defects; however, although not specifically studied, there are only
rare accounts of such lens abnormalities as cataract.36'37 Preliminary examinations of eyes from chimeric FGFR1 gene knockout mice (Dr. J. Rossant, personal communication, 1996) and of the FGFR3 homozygous mutant (Dr. C. X. Deng, personal communication, 1996) similarly suggest that eye development proceeds normally, but detailed studies of
lens differentiation have not yet been conducted.
In contrast, inhibition of all FGFR signaling in
transgenic mice using a dominant-negative strategy produced significant lens abnormalities. 1 3 1 4 Although these results do not exclude the possibility
of individual receptors having different functional
1697
roles in lens development, they do suggest there
may be some redundancy of FGFR function.
The four FGFRs expressed in the lens have different affinities for FGF family members. FGFR1 binds
FGF-1 and FGF-2 equally and with high affinity, but
binds FGF-4 with a lower affinity.38"40 FGFR2 (bek)
binds FGF-1, FGF-2, and FGF-4 with similar high affinity, but does not bind FGF-5 or FGF-738'41; KGFR binds
FGF-7 and FGF-1 with equal high affinity, but binds
FGF-2 with a much lower affinity.42'43 FGFR3 binds
FGF-1 with high affinity and FGF-2 with low affinity,44
and it has been shown that the mitogenic effects of
FGF-1 and FGF-4, mediated by FGFR3, are 10-fold
higher than the effect of FGF-2, whereas FGF-5 elicits
no response. 44 FGF-1 and FGF-2 can induce lens cell
proliferation, migration, and fiber differentiation in
vitro, but FGF-2 is more potent than FGF-1.2 Findings
in preliminary studies in which FGFs are overexpressed in lens fibers of transgenic mice indicate that
FGF-3, FGF-4, FGF-5, FGF-7, and FGF-8 also elicit fiberdifferentiation-like responses in lens epithelium 4 " 7 but
their relative potency in eliciting responses in lens
cells is unknown.
To date, expression of 5 of the 10 FGF family
members (FGF-1, FGF-2, FGF-3, FGF-5, and FGF-7)
has been detected in the developing eye. FGF-1 and
FGF-2 mRNA and protein are expressed widely in the
eye 2145 " 48 and are detected in the ocular media that
bathe the lens, with more of both forms in the vitreous
humor than in the aqueous humor. 2 ' 49 FGF-3, FGF-5,
and FGF-7 have more restricted expression patterns.
FGF-3 is expressed in the neural retina during development but is restricted to the peripheral neural retina at birth; it declines after birth and is undetectable
in the mature retina. 50 FGF-5 expression appears to
be restricted to the neural retina and pigmented epithelium, 5152 and FGF-7 is expressed in periocular mesenchyme during embryonic development 53 ; but in the
mature eye, expression appears to be restricted to the
cornea and conjunctiva.54'55 Clearly, cellular responses
to FGF in various regions of the lens in situ will depend
not only on the types and abundance of FGFRs present
but also on the types and abundance of FGF present.
Such responses may also depend on the capacity of
the FGFRs present to form particular homo- and heterodimer combinations on ligand binding (see introduction).
The current results establish that FGFRs are expressed in diverse temporal and spatial patterns during lens morphogenesis, differentiation, and growth
and highlight the complexity of the regulation of lens
cell behavior by FGF. There are many potential ways
in which the responses of lens cells to FGF may be
regulated in situ.2 Although the availability of ligands
in the extracellular milieu is a major consideration,
the current findings suggest that the regulation of
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 08/03/2017
1698
Investigative Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9
the spatiotemporal expression patterns of the various
FGFRs, with their capacity to form homodimers and
heterodimers, as well as their differing ligand binding
and signaling properties, may also play an important
role in regulating anteroposterior patterns of cell behavior in the lens.
Key Words
12.
13.
14.
fiber differentiation, fibroblast growth factor, fibroblast
growth factor receptor, lens development
Acknowledgments
The authors thank Dr. M. Bedford and Dr. P. Lonai (Weizmann Institute of Science, Rehovot, Israel) for the murine
bek and KGFR cDNAs, Dr. L. T. Williams (University of California, San Francisco, CA, USA) for the FGFR3 cDNA, Roland Smith for technical assistance with photography, and
staff of The University of Sydney Electron Microscope Unit
for the use of their image analysis facilities.
15.
16.
17.
References
1. Yamaguchi TP, RossantJ. Fibroblast growth factors in
mammalian development. Curr Opin Genet Dev.
1995; 5:485-491.
2. Chamberlain CG, McAvoy JW. Fibre differentiation
and polarity in the mammalian lens: A key role for
FGF. Prog Retina Eye Res. 1996; (In press).
3. McAvoy JW, Chamberlain CG. Fibroblast growth factor (FGF) induces different responses in lens epithelial cells depending on its concentration. Development.
1989; 107:221-228.
4. Chepelinsky AB, Robinson ML, Ash J, Parker-Wilson
DM, Ohtaka-Maruyama C, Overbeek, PA. Secreted
FGF-3 induces cell cycle withdrawal and differentiation of lens epithelia in transgenic mice. ARVO Abstracts. Invest Ophthalviol Vis Sci. 1996;37:S924.
5. Lovicu FJ, Srinivasan Y, Overbeek PA. Developmental
changes induced by lens-specific expression of different FGFs in transgenic mice. ICER Abracts. Exp Eye
Res. 1997;63(Suppl):S16.
6. Robinson ML, Overbeek PA, Verran DJ, et al. Extracellular FGF-1 acts as a lens differentiation factor in
transgenic mice. Development. 1995;'121:505-514.
7. Srinivasan Y, Overbeek PA. Expression of lens specific
fibroblast growth factor (FGF5) causes altered epithelial cell morphology and degeneration of fiber cells.
ARVO Abstracts. Invest Ophthalmol Vis Sci. 1996;
37:S924.
8. Johnson DE, Williams LT. Structural and functional
diversity in the FGF receptor multigene family. Adv
Cancel-Res. 1993; 60:1-41.
9. Klagsbrun M. Mediators of angiogenesis: The biological significance of basic fibroblast growth factor
(bFGF)-heparin and heparan sulfate interactions.
Semin Cancer Biol. 1992;3:81-87.
10. McKeehan WL, Kan M. Heparan sulfate fibroblast
growdi factor receptor complex: Structure-function
relationships. Mol Reprod Dev. 1994;39:69-81.
11. Jaye M, Schlessinger J, Dionne CA. Fibroblast growth
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
factor receptor tyrosine kinases: Molecular analysis
and signal transduction. Biochim Biophys Ada. 1992;
1135:185-199.
Schlessinger J, Ullrich A. Growth factor signaling by
receptor tyrosine kinases. Neuron. 1992;9:383-391.
Chow RL, Roux GD, Roghani M, et al. FGF suppresses
apoptosis and induces differentiation of fibre cells in
the mouse lens. Development. 1995; 121:4383-4393.
Robinson ML, MacMillan-Crow LA, Thompson JA,
Overbeek PA. Expression of a truncated FGF receptor
results in defective lens development in transgenic
mice. Development. 1995; 121:3959-3967.
de Iongh RU, Lovicu FJ, Hanneken A, Baird A, McAvoy JW. FGF receptor-1 (fig) expression is correlated
with fibre differentiation during rat lens morphogenesis and growth. DevDynam. 1996;206:412-426.
Orr-Urtreger A, Bedford MT, Burakova T, et al. Developmental localization of die splicing alternatives
of the fibroblast growth factor receptor-2. Dev Biol.
1993; 158:475-486.
Peters K, Ornitz D, Werner S, Williams LT. Unique
expression pattern of the FGF receptor 3 gene during
mouse organogenesis. Dev Biol. 1993; 155:423-430.
Yan G, McBride G, McKeeban WL. Exon skipping
causes alteration of the COOH-terminus and deletion
of the phospholipase-C-gamma 1 interaction site in
the FGF receptor 2 kinase in normal prostate epithelial cells. Biochem Biophys Res Commun. 1993; 194:512518.
Chellaiah AT, McEwan DG, Werner S, Xu J, Omitz
DM. Alternative splicing in immunoglobulin-like domain III creates a receptor highly specific for acidic
FGF/FGF-1./.BioZ Chem. 1994;269:11620-11627.
Simmons DM, Arriza JL, Swanson LW. A complete
protocol of in situ hybridization of messenger RNA in
brain and tissues with radiolabelled single-stranded
RNA probes. /Histotechnol. 1989; 12:169-181.
Lovicu FJ, de Iongh R, McAvoy JW. Expression of FGF1 and FGF-2 mRNA during lens morphogenesis, differentiation and growth. Curr Eye Res. 1997; 16:222223.
Wilkie AOM, Morris-Kay GM, Jones EY, Health JK.
Functions of fibroblast growth factors and their receptors. Curr Biol. 1995;5:500-507.
Bassnett S. The fate of the golgi apparatus and the
endoplasmic reticulum during lens fiber differentiation. Invest Ophthalmol Vis Sci. 1995;36:1793-1803.
Vainikka S, Partanen J, Bellosta P, et al. Fibroblast
growth factor receptor-4 shows novel features in genomic structure, ligand binding and signal transduction.
EMBOJ. 1992; 11:4273-4280.
WangJK, Gao G, Goldfarb M. Fibroblast growth factor
receptors have different signalling and mitogenic potentials. Mol Cell Biol. 1994; 14:181-188.
Cenedella RJ. Aging and rates of lens cell differentiation in vivo, measured by a chemical approach. Invest
Ophthalmol Vis Sci. 1989; 30:575-579.
Lovicu FJ, McAvoy JW. The age of rats affects the
response of lens epithelial explants to fibroblast
growth factor: An ultrastructural analysis. Invest Ophthalmol Vis Sci. 1992;33:2629-2278.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 08/03/2017
FGF Receptor Expression in Lens Development
28. Richardson NA, McAvoy JW, Chamberlain CG. Age
of rats affects response of lens epithelial explants to
fibroblast growth factor. Exp Eye Res. 1992; 55:649656.
29. Amaya E, Musci TJ, Kirschner M. Expression of a dominant negative mutant of the FGF receptor disrupts
mesoderm formation in Xenopus embryos. Cell.
1991; 66:257-270.
30. Ueno H, Gunn M, Dell K, Tseng A, Williams L. A
truncated form of fibroblast growth factor 1 inhibits
signal transduction by multiple types of fibroblast
growth factor receptor. / Biol Chem. 1992; 267:14701476.
31. Muenke M, Schell U. Fibroblast-growth-factor receptor mutations in human skeletal disorders. Trends
Genet. 1995; 11:308-313.
32. Jabs EW, Li X, Scott AF, et al. Jackson-Weiss and
Crouzon syndromes are allelic with mutations in fibroblast growth factor receptor 2. Nat Genet. 1994;
8:275-279.
33. Deng C, Wynshaw-Boris A, Zhou F, Kuo A, Leder
P. Fibroblast growth factor receptor 3 is a negative
regulator of bone growth. Cell. 1996;84:911-921.
34. Deng C, Wynshaw-Boris A, Shen MM, Daugherty C,
Ornitz DM, Leder P. Murine FGFR-1 is required for
early postimplantation growth and axial organization.
Genes Dev. 1994;8:3045-3057.
35. Yamaguchi TP, Harpal K, Henkemeyer M, Rossant J.
fgfr-1 is required for embryonic growth and mesodermal patterning during mouse gastrulation. Genes Dev.
1994;8:3032-3044.
36. Howell SC. Craniostenosis. Report of 22 cases. Am J
Ophthalmol. 1954; 37:359-367.
37. Menashe Y, Ben Baruch G, Rabinovitch O, Shalev Y,
Katzenlson MBM, Shalev E. Exophthalmus—prenatal
ultrasonic features for diagnosis of Crouzon syndrome. PrenatDiagn. 1989;9:805-808.
38. Dionne CA, Crumley G, Bellot F, et al. Cloning and
expression of two distinct high-affinity receptors crossreacting with acidic and basic fibroblast growth factors. EMBOj. 1990; 9:2685-2692.
39. Johnson DE, Lee PL, Lu J, Williams LT. Diverse forms
of a receptor for acidic and basic fibroblast growth
factors. Mol Cell Biol. 1990; 10:4728-4736.
40. Mansukhani A, Dell'Era P, Moscatelli D, Kornbluth S,
Hanafusa H, Basilico C. Characterization of the murine BEKfibroblastgrowth factor (FGF) receptor: Activation by three members of the FGF family and requirement for heparin. Proc Nail Acad Sci USA.
1992;89:3305-3309.
41. Mansukhani A, Moscatelli D, Talarico D, Levytska V,
Basilico C. A murine fibroblast growth factor (FGF)
receptor expressed in CHO cells is activated by basic
FGF and Kaposi FGF. Proc Natl Acad Sci USA.
1990;87:4378-4382
1699
42. Bottaro DP, Rubin JS, Ron D, Finch PW, Florio C,
Aaronson SA. Characterization of the receptor for keratinocyte growth factor—evidence for multiple fibroblast growth factors. /Biol Chem. 1990;265:1276712770.
43. Miki T, Bottaro DP, Fleming TP, et al. Determination
of the ligand-binding specificity by alternative splicing.
Two distinct growth factor receptors encoded by a
single gene. Proc Natl Acad Sci USA. 1992; 89:246-250.
44. Ornitz DM, Leder P. Ligand specificity and heparin
dependence of fibroblast growth factor receptors 1
and 3. J Biol Chem. 1992; 267:16305-16311.
45. de Iongh R, McAvoy JW. Distribution of acidic and
basic fibroblast growth factors (FGF) in the foetal rat
eye: Implications for lens development. Growth Factors.
1992;6:159-177.
46. de Iongh R, McAvoy JW. Spatio-temporal distribution
of acidic and basic FGF indicates a role for FGF in rat
lens morphogenesis. Dev Dynam. 1993; 198:190-202.
47. Noji S, Matsuo T, Koyama E, et al. Expression pattern
of acidic and basic fibroblast growth factor genes in
adult rat eyes. Biochem Biophys Res Commun. 1990;
168:343-349.
48. Lovicu FJ, McAvoy JW. Immunolocalization of acidic
FGF, basic FGF and heparan sulfate proteoglycan in
the rat lens: Implications for lens polarity and growth
patterns. Invest Ophthalmol Vis Sci. 1993; 34:3355-3365.
49. Schulz M, Chamberlain CG, de Iongh RU, McAvoy JW.
Acidic and basic FGF in ocular media: Implications for
lens polarity and growth patterns. Development. 1993;
118:117-126.
50. Wilkinson DG, Bhatt S, Kennedy AP. Expression of
the FGF-related proto-oncogene int-2 suggests multiple roles in fetal development. Development. 1989;
105:131-136.
51. Bost LM, Aotaki-Keen AE, Hjelmeland LM. Coexpression of FGF-5 and bFGF by the retinal pigment
epithelium in vitro. Exp Eye Res. 1992; 55:727-34.
52. Kitaoka T, Aotaki-Keen AE, Hjelmeland LM. Distribution of FGF-5 in the rhesus macaque retina. Invest
Ophthalmol Vis Sci. 1994;35:3189-3198.
53. Finch PW, Cunha GR, Rubin JS, WongJ, Ron D. Pattern of keratinocyte growth factor and keratinocyte
growth factor receptor expression during mouse fetal
development suggests a role in mediating morphogenetic mesenchymal-epithelial interactions. Dev Dynam. 1995; 203:223-240.
54. Wilson SE, Walker JW, Chwang EL, He YG. Hepatocyte growth factor, keratinocyte growth factor, their
receptors, fibroblast growth factor receptor-2, and the
cells of die
cornea.
Invest Ophthalmol Vis Sci.
1993; 34:2544-2561.
55. Li DQ, Tseng SC. Three patterns of cytokine expression potentially involved in epithelial-fibroblast interactions of human ocular surface. / Cell Physiol.
1995; 163:61-79.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 08/03/2017